My Projects

An overview of recent and ongoing scientific Projects I've been working on.

We developed a Data Science Competition for ecologists aiming to predict where and which individual trees are distributed across the landscape using hiperdimensional remote sensing data.

Data Science Competition in Ecology

We developed a machine learning method to predict average chemical and phisical quantities for million of trees using a combination of field and remote sensing data. Ah, did I say our products are openly usable?

Rethinking Remote Sensing Unit

We developed a fully deterministic process based model of forest ecosystems growth: with this we can forecast Carbon Exchange, Forest Growth, and Allocation behaviour.

3D CMCC Forest Ecosystem Model

Cross-Scale Macrosystems: Functional Pseudo-Species Patterns

Cross-Scale Macrosystems: Tree Species Patterns

https://www.ecodse.org

Hierarchical multiple Traits Regression

Remote sensing models are generally built on pixel to response relationship. But are sure we picked the right pixels? To address the uncertainty associated to pixel choise in Remote Sensing Regressions we are developing a Multiple Instance method to be paired to the model used to estimate response given such uncertainty.

Joint Multiple Instance Regression